Glass-Filled Polybutylene Terephthalate: What the Label Percentage Does to Shrinkage

The mold shrinkage of Polybutylene Terephthalate (PBT) is not a single material constant; it is a process- and geometry-dependent response measured on standardized plaques. Glass-filled PBT compounds are labeled by wt% glass reinforcement, but the volumetric glass fraction, residual fibre length distribution, and fibre orientation through the cavity thickness determine how much restraint is applied to the semicrystalline PBT matrix. Datasheet shrinkage figures are generated according to ISO 294-4:2018 or ASTM D955-21 using 60 mm × 60 mm × 2 mm plaques and conditioning at 23 °C and 50% relative humidity. The label percentage is a nominal formulation target, not an explicit predictor of final part dimensions.

Why Does Glass Fibre Content Reduce Mold Shrinkage in PBT?

Unfilled PBT exhibits flow-direction mold shrinkage typically ranging from 1.5–2.0% and transverse shrinkage from 1.8–2.2%. The addition of E-glass fibre reduces both values, but the reduction is not linear with the label percentage because shrinkage restraint depends on volume fraction and orientation. E-glass has a coefficient of linear thermal expansion near 5 × 10^-6 K^-1, while the PBT matrix expands and contracts at 80–110 × 10^-6 K^-1. A compound labeled 30 wt% glass is not 30 vol% glass. Using a PBT matrix density of 1.31 g/cm³ and an E-glass density of 2.54–2.62 g/cm³, 30 wt% fibre converts to approximately 18 vol%. The volumetric relationship follows:

φ_v = (w_f / ρ_f) / ((w_f / ρ_f) + (w_m / ρ_m))

In the flow direction, oriented fibres mechanically restrict linear contraction of the surrounding PBT matrix. In the transverse direction, where fewer fibres are aligned, matrix shrinkage remains higher. This is why a single label percentage cannot describe the dimensional behavior of a glass-filled PBT part in two orthogonal directions.

In a 30 wt% glass-filled PBT, the flow-direction shrinkage may be 0.2–0.5% while the transverse direction remains 0.5–0.9%. The difference between flow and transverse shrinkage is a more useful processing parameter than the single label percentage because it drives warpage in flat parts with ribs, bosses, or gate asymmetry. Fibre orientation through a 2 mm plaque thickness follows a skin–core–skin profile. The skin layers contain fibres aligned predominantly in the flow direction due to extensional fountain flow and rapid freezing at the cavity wall, while the core layer contains fibres oriented more transversely under shear flow. This layered orientation amplifies in-plane shrinkage anisotropy even when the part is labeled 30 wt% glass-filled.

When Gate Freeze Occurs Before Packing Compensation Is Complete

Shrinkage in glass-filled PBT is strongly sensitive to the timing of gate freeze relative to packing pressure decay. If the gate freezes before cavity pressure decays to atmospheric pressure, the cavity cannot receive additional melt to compensate for thermal contraction. This produces higher mass variation and elevated shrinkage along the flow path. In cold-runner tools with a 2 mm wall, gate freeze times for a 30 wt% glass-filled PBT at melt temperature 260 °C and mold temperature 80 °C can range from 1.5–3.0 s. Raising hold pressure from 40 MPa to 80 MPa increases cavity pressure at freeze and typically reduces flow-direction shrinkage by 0.1–0.3 percentage points. These values shift when mold temperature falls below 60 °C because the solidified skin thickens, the effective flow channel narrows, and pressure transmission to the cavity extremity deteriorates.

Injection molding of glass-filled PBT requires melt temperatures between 250 °C and 270 °C and mold temperatures between 60 °C and 100 °C. Lower mold temperatures reduce cycle time but produce higher orientation in the skin and less matrix crystallinity near the surface, which can increase post-mold instability. Higher mold temperatures improve fibre wetting and pressure transmission but extend cycle time and can raise transverse crystallinity enough to increase long-term shrinkage in thick sections. Glass-filled PBT also has higher melt viscosity than unfilled PBT. Thin-walled parts with 1.5 mm wall stock typically require injection pressures of 80–120 MPa. Clamp force calculations for 30 wt% glass-filled PBT commonly use 0.5–0.8 tonnes per cm² of projected area, depending on flow length and wall thickness. Insufficient clamp force creates flash, but flash also reduces packing effectiveness and distorts shrinkage measurements.

Before any melt-state shrinkage assessment can be valid, the moisture content of PBT must be reduced below 0.02 wt%. Water above this threshold at melt temperatures above 250 °C triggers autocatalytic hydrolytic chain scission that lowers molecular weight and melt viscosity, producing short shots, splay, and inconsistent packing. Drying in desiccant dryers at 120–140 °C for 4 h with a dew point below -30 °C is standard practice; moisture measurement can follow ISO 15512:2019. Glass-filled PBT is not a general-purpose material that tolerates open-air hopper loading under humid conditions. At relative humidity above 60%, regrind and virgin pellets can regain surface moisture within 30–60 min if the hopper is not sealed or purged with dry air. The resulting lot-to-lot viscosity drift changes gate freeze time and cavity pressure response even when the label percentage is unchanged.

Thermal Expansion, Crystallinity, and the Role of Post-Mold Conditioning

Post-mold shrinkage continues after ejection because PBT crystallizes slowly at low ambient temperatures. Parts conditioned at 23 °C and 50% relative humidity can show an additional 0.02–0.1% dimensional change relative to immediate after-ejection measurements. This post-mold change is partly thermal contraction, partly continued secondary crystallization, and partly moisture uptake. PBT has a glass transition temperature near 45–60 °C, so at upper service temperatures approaching 80 °C, further crystallization can occur. Annealing glass-filled PBT at 120 °C for 2 h can reduce subsequent long-term creep strain but may add 0.05–0.3% additional shrinkage depending on fibre orientation and initial cooling rate. The lower coefficient of linear thermal expansion in the flow direction means that anisotropic parts annealed at elevated temperature often show differential dimensional change between flow and transverse directions. For this reason, post-mold conditioning is a dimensional stability variable, not merely a laboratory convenience.

On compounding lines, a given glass weight percentage can produce different shrinkage after injection molding because the residual fibre length after extrusion determines aspect ratio. Initial chopped strands of 4.5 mm are fed downstream after polymer melting on a segmented twin-screw extruder with an L/D ratio of 40:1; kneading blocks and screw speeds of 300–600 rpm can reduce the number-average fibre length to 200–400 μm. Lower screw speeds and low-shear screw profiles preserve longer fibres and produce lower flow-direction shrinkage, but they also increase the risk of fibre bundles, poor dispersion, and large numerical differences between flow and transverse shrinkage. Worn screw elements increase fibre breakage, causing shrinkage to drift upward over a production campaign without any change in the label percentage. Two compounds labeled 30 wt% glass-filled PBT may therefore exhibit measurably different mold shrinkage on the same production mold if their fibre length retention, sizing chemistry, and stabilizer packages differ.

In multi-cavity tools, the apparent shrinkage of a 30 wt% glass-filled PBT can differ between cavities by 0.05–0.15 percentage points solely due to shear heating, runner imbalance, or gate diameter variations of ±0.1 mm. In hot-runner systems, melt temperature differences larger than ±3 °C across cavities are sufficient to shift local crystallization rates and shrinkage. Weld lines formed downstream of ribs or bosses interrupt flow orientation and create local shrinkage anisotropy that can exceed 0.2 percentage points relative to adjacent bulk material. Because the label percentage does not capture these tooling effects, dimensional capability studies on production tools remain necessary even when material datasheet values are available. Tolerance stack-ups for glass-filled PBT require separate allowance for flow-direction shrinkage, transverse shrinkage, post-mold crystallization shrinkage, and cavity-to-cavity pressure imbalance.